How does Quine argue to the indeterminacy of translation? It has always been hard to tell, but one construction goes like this: A fact is determinate only if it is empirically determinable, for empirical evidence under-determines truth about the trans-empirical; theory is underdetermined by data. Therefore, facts about the unobservable, facts of theory, are indeterminate. A fact about the meaning of a person's sentence is only determinate if it is empirically determinable. The meaning of a person's sentence is empirically determinable only if it is empirically determinable under what conditions he assents to it. The meaning of a person's sentence about something unobservable is determinable only if it is determinable what unobservable events are occurring whenever he assents to a sentence supposedly about the unobservable. It is not empirically determinable what is going on in the unobservable realm whenever someone assents to a sentence supposedly about the unobservable. Therefore it is undeterminable what a person's theory sentence means. Therefore what it means is indeterminate.